Electronic Design

Los Alamos Readies Spaceborne Supercomputer

Research laboratories here on Earth can take advantage of supercomputers as they use their awesome processing power to simulate nuclear blasts, chart weather patterns, and map the human genome. But due to space and power limitations, researchers in orbit don’t have those advantages—at least not yet. Los Alamos National Laboratory is developing a space payload that dramatically increases on-orbit computational capabilities.

The payload will process more than 1000 Goperations/s for software-defined radio (SDR) functions in space. SDR has military applications in supporting tactical communications as well as potential commercial broadcasting applications. Despite this processing power, the payload will weigh only 40 lb and consume 80 W, thanks to the 90-nm Virtex-4 platform FPGA from Xilinx Inc. The payload also will include Atmel’s AT697 RadHard SPARC processor and BAE Systems’ chalcogenide C-RAM.

“Our sensors in the Global Positioning System and Defense Satellite Program platforms have been severely constrained by the data downlinks available,” said Mark Hodgson, Los Alamos’ portfolio manager for the Space Nuclear Explosion Monitoring program. “This new reprogrammable, supercomputing-payload technology enables our science staff to use in space the algorithms and methods previously only possible in ground-based mainframe computers.”

The National Nuclear Security Administration’s Office of Nonproliferation Research and Development and the U.S. Department of Defense are sponsoring the payload’s development. Researchers say that the increase in processing power will help them discriminate between signals related to nuclear explosions and natural and man-made background signals.

Los Alamos National Laboratory
www.lanl.gov

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish